Waist circumference thresholds for the prediction of cardiometabolic risk: is measurement site important?

Waist circumference (WC) is frequently measured in clinical and research settings. Although measurement protocols may differ considerably, a single set of sex-specific cut-points are typically used to denote elevated risk. The purpose of this study was to determine whether the appropriate WC thresholds for identifying cardiometabolic risk vary according to anatomical measurement site.

WC was measured at four common sites in 520 community-dwelling adults (20-66 years): superior border of the iliac crest, midpoint between the iliac crest and the lowest rib, umbilicus and minimal waist. Resting blood pressures and fasting levels of total cholesterol, high-density lipoprotein-cholesterol, triglycerides and glucose were measured. The sensitivity and specificity of current WC thresholds (M: >102 cm/40 inch; W: >88 cm/35 inch) for detecting abnormal risk factor levels were calculated for each WC measurement site, and receiver operating characteristic (ROC) curves were used to select optimal thresholds for identifying individuals with risk factor clustering (>or=2 risk factors).

The area under the ROC curve for WC measured at each anatomical site was similar for each risk factor and for the presence of risk factor clustering. However, WC >88/102 cm at the umbilicus showed the greatest sensitivity for all outcomes, whereas measurements at the minimal waist had the best specificity in this sample. The sensitivity of WC >88/102 cm for detecting >or=2 risk factors ranged from 75 to 89% in women and from 48 to 59% in men, and specificity ranged from 52 to 79% in women and from 77 to 88% in men, across measurement sites.

At present, recommended WC thresholds may not have the same clinical utility at all anatomical locations of WC measurement.